Abstract: SA-PO414

Prediction of Urinary Creatinine Excretion and of Renal Function in CKD Patients from Body Composition Analysis

Session Information

Category: Chronic Kidney Disease (Non-Dialysis)

  • 302 CKD: Estimating Equations, Incidence, Prevalence, Special Populations

Author

  • Donadio, Carlo, University of Pisa, Pisa, Italy
Background

The utility of creatinine clearance (CCr) to measure renal function is greatly reduced by its low accuracy and by the need for a timed collection of urine. Different formulas have been proposed to predict CCr or GFR from serum creatinine (PCr) and anthropometric data. However, the inaccuracy in the estimate of urinary creatinine excretion (UCr) is probably the major cause of error of prediction formulas.
The aim of this study was to evaluate the possibility to obtain a more accurate prediction of UCr, and hence of CCr from the measure of body cell mass (BCM). BCM is the body compartment which contains muscle mass, where creatinine is generated.

Methods

Two-hundred and eleven CKD patients (115 women), aged 17-81 years, PCr 0.5-14.4 mg/dL. Glomerular filtration rate (GFR) was measured in all patients as the renal clearance of 99mTc-DTPA. The other examined parameters were: PCr; UCr (urine collection 24 hrs); CCr measured as UCr x V / PCr (m-CCr); CCr predicted by Cockcroft and Gault formula (C&G-CCr). BCM was measured using a single frequency tetrapolar impedance analyzer.

Results

GFR was 1.3-147 mL/min (mean 55.1), mCCr 3.7-135 mL/min (mean 61.2), 24h-UCr 416-2102 mg (mean 1075), BCM 12.2-45.2 kg (mean 23.7). A strict linear correlation was found between 24h-UCr and BCM (r=0.780), which was closer than that between 24h-UCr and BW (r=0.626). Multiple linear regression (MR) modeling for urinary creatinine excretion indicated that UCr was determined positively by BCM, body weight, male gender and height, and negatively by age and PCr. The most relevant determinants of UCr were BCM, age, and PCr. UCr predicted from MR equation (MR-UCr) ranged 378-1856 mg (mean 1075) quite similar to measured 24h-UCr. MR-BCM-CCr, that is CCr predicted from MR-UCr and PCr, was very similar to measured CCr (61.5±32.0 VS 61.2±32.1 mL/min, NS), with a closer correlation than C&G CCr (r= 0.940 VS 0.857) and a lower prediction error (11.1 VS 13.0 mL/min). The concordance between MR-BCM-CCr and mCCr was better than that of C&G CCr .

Conclusion

In CKD patients at the different stages of GFR impairment, urinary creatinine excretion can be more accurately estimated, and creatinine clearance can be predicted from the value of body cell mass combined with other anthropometric data and with serum creatinine.

Funding

  • Government Support - Non-U.S.